A5027608625- Bioinformatics and Genomic Networks
- Gene expression and cancer classification
- Computational Drug Discovery Methods
- Machine Learning in Bioinformatics
- RNA Research and Splicing
- Cancer Genomics and Diagnostics
- Gene Regulatory Network Analysis
- RNA modifications and cancer
- Biomedical Text Mining and Ontologies
- Genomic variations and chromosomal abnormalities
- Genetics, Bioinformatics, and Biomedical Research
- Cancer Immunotherapy and Biomarkers
- Hepatocellular Carcinoma Treatment and Prognosis
- Chromosomal and Genetic Variations
- Pharmacogenetics and Drug Metabolism
- Cancer Treatment and Pharmacology
- Monoclonal and Polyclonal Antibodies Research
- Protein Structure and Dynamics
- Genomics and Phylogenetic Studies
- Cancer Mechanisms and Therapy
- Cancer-related molecular mechanisms research
- HER2/EGFR in Cancer Research
- RNA and protein synthesis mechanisms
- Ferroptosis and cancer prognosis
- RNA regulation and disease
Incheon National University
2017-2025
OHSU Knight Cancer Institute
2019
Cedars-Sinai Medical Center
2019
Providence Portland Medical Center
2019
University of California, Los Angeles
2014-2018
UCLA Health
2015
Yonsei University
2008-2014
Gachon University
2011-2012
Sungkyunkwan University
2011-2012
University of Ulsan
2012
Tumor metastasis remains the major cause of cancer-related death, but its molecular basis is still not well understood. Here we uncovered a splicing-mediated pathway that essential for breast cancer metastasis. We show RNA-binding protein heterogeneous nuclear ribonucleoprotein M (hnRNPM) promotes by activating switch alternative splicing occurs during epithelial–mesenchymal transition (EMT). Genome-wide deep sequencing analysis suggests hnRNPM potentiates TGFβ signaling and identifies CD44...
Adenosine deaminases acting on RNA (ADARs) are the primary factors underlying adenosine to inosine (A-to-I) editing in metazoans. Here we report first global study of ADAR1-RNA interaction human cells using CLIP-seq. A large number CLIP sites observed Alu repeats, consistent with ADAR1's function editing. Surprisingly, thousands other located non-Alu regions, revealing functional and biophysical targets ADAR1 regulation alternative 3' UTR usage miRNA biogenesis. We observe that binding UTRs...
Predicting the effect of drug-drug interactions (DDIs) precisely is important for safer and more effective drug co-prescription. Many computational approaches to predict DDIs have been proposed, with aim reducing effort identifying these in vivo or vitro, but room remains improvement prediction performance.In this study, we propose a novel deep learning model accurately.. The proposed uses autoencoders feed-forward network that are trained using structural similarity profiles (SSP), Gene...
Although major genetic networks controlling early liver specification and morphogenesis are known, the mechanisms responsible for postnatal hepatic maturation poorly understood. Here we employ global analyses of mouse transcriptome to demonstrate that remodelling is accompanied by large-scale transcriptional post-transcriptional transitions cell-type-specific temporally coordinated. Combining detailed expression with gain- loss-of-function studies, identify epithelial splicing regulatory...
Accurate prediction of protein-ligand binding affinity is important for lowering the overall cost drug discovery in structure-based design. For accurate predictions, many classical scoring functions and machine learning-based methods have been developed. However, these techniques tend to limitations, mainly resulting from a lack sufficient energy terms describe complex interactions between proteins ligands. Recent deep-learning can potentially solve this problem. search more efficient...
The epithelial–mesenchymal transition (EMT) is a fundamental developmental process that abnormally activated in cancer metastasis. Dynamic changes alternative splicing occur during EMT. ESRP1 and hnRNPM are regulators promote an epithelial program mesenchymal program, respectively. functional relationships between these factors the genome scale remain elusive. Comparing targets of revealed they coregulate set cassette exon events, with majority showing discordant regulation. Discordant...
Cancer is one of the most difficult diseases to treat owing drug resistance tumour cells. Recent studies have revealed that responses are closely associated with genomic alterations in cancer Numerous state-of-the-art machine learning models been developed for prediction using various data and diverse molecular information, but those methods ineffective predict response untrained drugs gene expression patterns, which known as cold-start problem. In this study, we present a novel deep neural...
Background The prognosis of cancer recurrence is an important research area in bioinformatics and challenging due to the small sample sizes compared vast number genes. There have been several attempts predict recurrence. Most studies employed a supervised approach, which uses only few labeled samples. Semi-supervised learning can be great alternative solve this problem. based on manifold assumptions reveal detailed roles identified genes Results In order recurrence, we proposed novel...
Background IgA neutralizes pathogens to prevent infection at mucosal sites. However, emerging evidence shows that contributes aggravating inflammation or dismantling antitumor immunity in human diseased liver. The aim of this study was elucidate the roles inflammation-induced intrahepatic inflammatory + monocytes development hepatocellular carcinoma (HCC). Methods Patient cohorts including steatohepatitis cohort (n=61) and HCC (n=271) were established. Patients’ surgical biopsy specimens...
The growing number and variety of genetic network datasets increases the feasibility understanding how drugs diseases are associated at molecular level. Properly selected features representations existing drug-disease associations can be used to infer novel indications drugs. To find new associations, we generated an integrative using combinations interactions, including protein-protein interactions gene regulatory datasets. Within this network, adjacencies drug-drug disease-disease were...
Accurate identification of prognostic biomarkers is an important yet challenging goal in bioinformatics. Many bioinformatics approaches have been proposed for this purpose, but there still room improvement. In paper, we propose a novel machine learning-based method more accurate biomarker genes and use them prediction cancer prognosis. The specifies the candidate gene module by graph learning using generative adversarial networks (GANs) model, scores PageRank algorithm. We applied to...
With the increasing incidence of breast cancer worldwide, in particular southeast Asia (including Korea), and common use anthracyclines adjuvant metastatic settings, occurrence Hepatitis B virus (HBV) reactivation may develop this patient population. The prophylactic antiviral agents patients result a reduced HBV exacerbation. purpose current study was to assess efficacy lamivudine reducing severity post-operative undergoing doxorubicin-containing chemotherapy. medical records...
Abstract Motivation Identification of genes that can be used to predict prognosis in patients with cancer is important it lead improved therapy, and also promote our understanding tumor progression on the molecular level. One common but fundamental problems render identification prognostic prediction outcomes difficult heterogeneity patient samples. Results To reduce effect sample heterogeneity, we clustered data samples using K-means algorithm applied modified PageRank functional...
Abstract Motivation: Diagnosis and prognosis of cancer understanding oncogenesis within the context biological pathways is one most important research areas in bioinformatics. Recently, there have been several attempts to integrate interactome transcriptome data identify subnetworks that provide limited interpretations known candidate genes, as well increase classification accuracy. However, these studies little information about detailed roles identified genes. Results: To more network, we...
Identification of cancer prognostic genes is important in that it can lead to accurate outcome prediction and better therapeutic trials for patients. Many computational approaches have been proposed achieve this goal; however, there room improvement. Recent developments deep learning techniques aid the identification more prediction, but one main problems adoption purpose data from patients too many dimensions, while number samples relatively small. In study, we propose a novel network-based...
Optimizing techniques for discovering molecular structures with desired properties is crucial in artificial intelligence (AI)-based drug discovery. Combining deep generative models reinforcement learning has emerged as an effective strategy generating molecules specific properties. Despite its potential, this approach ineffective exploring the vast chemical space and optimizing particular To overcome these limitations, we present Mol-AIR, a learning-based framework using adaptive intrinsic...
Predicting immune checkpoint inhibitor (ICI) response remains a significant challenge in cancer immunotherapy. Many existing approaches rely on differential gene expression analysis or predefined signatures, which may fail to capture the complex regulatory mechanisms underlying response. Network-based models attempt integrate biological interactions, but they often lack quantitative framework assess how individual genes contribute within pathways, limiting specificity and interpretability of...
Abstract Motivation: Accurate identification of genetic variants such as single-nucleotide polymorphisms (SNPs) or RNA editing sites from RNA-Seq reads is important, yet challenging, because it necessitates a very low false-positive rate in read mapping. Although many aligners are available, no single aligner was specifically developed tested an effective tool for SNP and prediction. Results: We present RASER, accurate with novel mapping schemes index tree structure that aims to reduce...
The purpose of this study is to validate the recently published Breast–Graded Prognostic Assessment (GPA) and propose a new prognostic model nomogram for patients with brain parenchymal metastases (BM) from breast cancer (BC). We retrospectively investigated 171 consecutive who received diagnosis BM BC during 2000–2008. appraised proposed Sperduto's BC-specific GPA in training cohort through Kaplan-Meier survival curve using log-rank test area under BC-GPA predicting overall at 1 year...
We propose a novel method that predicts binding of G-protein coupled receptors (GPCRs) and ligands. The proposed uses hub cycle structures ligands amino acid motif sequences GPCRs, rather than the 3D structure receptor or similarity experimental results show these new features can be effective in predicting GPCR-ligand (average area under curve [AUC] 0.944), because they are thought to include hidden properties good ligand-receptor binding. Using method, we were able identify ligand-GPCR...
This study aimed to compare the prognosis and characteristics of patients with advanced hepatocellular carcinoma treated first-line atezolizumab plus bevacizumab (AB) combination therapy hepatic artery infusion chemotherapy (HAIC). We retrospectively assessed 193 114 HAIC AB therapy, respectively, between January 2018 May 2023. The progression-free survival (PFS) was significantly superior that (p < 0.05), but there no significant difference in overall (OS). After propensity score...
Accurate prediction of cancer stage is important in that it enables more appropriate treatment for patients with cancer. Many measures or methods have been proposed accurate stage, but recently, machine learning, especially deep learning-based receiving increasing attention, mostly owing to their good accuracy many applications. Machine learning can be applied high throughput DNA mutation RNA expression data predict stage. However, because the number genes markers generally exceeds 10,000, a...
Machine learning may be a powerful approach to more accurate identification of genes that serve as prognosticators cancer outcomes using various types omics data. However, date, machine approaches have shown limited prediction accuracy for outcomes, primarily owing small sample numbers and relatively large number features. In this paper, we provide description GVES (Gene Vector Each Sample), proposed model can efficiently leveraged even with size, increase the prognostic value. GVES, an...